L2P-Norm Distance Twin Support Vector Machine

Abstract

A twin support vector machine (TWSVM) is an effective classifier, especially for binary data, which is defined by squared l2-norm distance in the objective function. Since squared l2-norm distance is susceptible to outliers, it is desirable to develop a revised TWSVM. In this paper, a new robust TWSVM via l2,p-norm formulations was proposed, because it suppress the influence of outliers better than l1-norm or squared l2-norm minimizations. However, the resulted objective is challenging to solve, because it is non-smooth and non-convex. As an important work, we systematically derive an efficient iterative algorithm to minimize the pth order of l2-norm distances. Theoretical support shows that the iterative algorithm is effective in the resolution to improve TWSVM via l2,p-norm instead of squared l2-norm distances. A large number of experiments show that l2,p-norm distances twin support vector machine can treat the noise data effectively and has a better accuracy.

Keywords

• TWSVM
• L2P-norm
• robustness